Background of the Invention
[0001] The present invention relates to a continuous speech recognition method and device
for automatically recognizing a plurality of concatenated words such as numericals
and for producing an output in accordance with the recognized content.
[0002] Speech recognition devices have been considered to be effective means for performing
man/machine communication. However, most of the devices which have been developed
so far have a disadvantage in that only isolated or discrete words can be recognized,
so that data input speed is very low. In order to solve the above problem, a continuous
speech pattern device which uses a two-level dynamic programming (to be referred to
as a two-level DP) algorithm is described in Patent Disclosure DE-A 26 10 439 In the
principle of this algorithm, pattern strings obtained by concatenating several reference
patterns in all possible orders are defined as reference pattern strings of a continuous
speech. An input pattern as a whole is mapped onto the reference pattern strings.
The number of reference patterns and their arrangement are determined to maximize
the overall similarity measure between the input pattern and the reference pattern
strings. Thus, speech recognition is performed. In practice maximization is achieved
by two stages of maximization; of individual words and of word strings. These maximizations
can be performed utilizing the DP algorithm.
[0003] The two-level DP algorithm will be described in detail below.
[0004] Let feature vectors α
i be
then, a speech pattern A is defined as a time series of α
i:
where I is the duration of the speech pattern A, and Q is the number of components
of the feature vectors. Thus, the speech pattern A is regarded as the input pattern.
[0005] Assume that N reference patterns B
n(n = 1, 2,..., N) are defined as a set of words to be recognized. Each reference pattern
B
n has J
n feature vectors as follows:
where the feature vector
is a similar vector to the feature vector α
i, as follows:
[0006] The partial pattern of the input pattern A which has a starting point ℓ and an endpoint
m on the time base i can be expressed as follows:
for 1 ≤ ℓ < m < I
[0007] Between the partial pattern A
(ℓ m) and the reference pattern B
n, a function j(i) which establishes a correspondence between a time base i of the
input pattern and a time base j of the reference pattern is optimally determined.
Partial matching is performed wherein a maximum value S(A,
m)'B
n) of the sum of similarity measures s(α
i,
(to be referred to as s (i, j)) between vectors which are defined by i and j(i) is
computed by the DP algorithm. In the first stage, a partial similarity measure S<ℓ,
m> as the maximum value of S(A
(ℓ, m), B
n) is determined for n which is computed by sequentially changing the starting point
1 and the endpoint m, and a partial determination result n<ℓ, m> for providing the
maximum value is also determined. Overall matching is performed at the second stage
wherein the number Y of words included in the input pattern and boundaries ℓ
(1), ℓ
(2),..., ℓ
(Y-1) which number (Y - 1) are optimally determined, and wherein the number Y of words
and boundaries ℓ̂
(1), ℓ̂
(2),..., ℓ̂
(Ŷ -1) are obtained to maximize the sum of the partial similarity measures during continuous
and nonoverlapping duration. The sum is given by the following relation:
The boundaries ℓ̂
(1), ℓ̂
(2),..., ℓ̂
(Ŷ - 1) and the partial determination result
n<1, m> determine n<1, ℓ̂
(1)>,
[0008] The definition of the similarity measure is given by a function which maps the time
base j of a reference pattern B and the time base i of the input pattern A in order
to correct deviation between the time bases of the input pattern A given by relation
(2) and the reference pattern B given by relation (3) as follows:
Assume that the similarity measure s(i, j) is exemplified by the following relation:
The similarity measure between the input pattern A and the reference pattern B is
given as follows:
It is impossible to obtain a maximum value of relation (9) by computing all the possibilities
for j = j(i). Instead, the DP algorithm is utilized as follows. Let the initial conditions
be:
g(I, J) is computed in a range of i = 2 to I and j = 1 to J by the following recursive
relation:
Therefore, S(A, B) of relation (9) is given by:
The deviation of the time base in practice may not exceed 50% in practice, so that
a hatched region bounded by lines 11 and 12 and a line 15 indicated by "i = j" in
Fig. 1 need only be considered. Therefore, recursive relation (11) need only be applied
in the range:
The above hatched region is called an adjustment window.
[0009] The partial similarity measure of the endpoint m in the range indicated by reference
numeral 14 in Fig. 1 is obtained in correspondence with one starting point ℓ. The
hatched region in Fig. 1 is defined by all the computations "(2 * r + 1) * J
n" for one starting point.
[0010] When relation (13) is used as a condition for the alignment range of time bases i
and j, the total computation C
1 by the DP algorithm for the similarity measure s(i, j) is approximated as follows,
even if only the partial similarity measure "S<ℓ, m>" is to be obtained in the first
stage:
where I is the duration of the input pattern, N is the number of reference patterns,
and J is the average duration of the reference patterns. For the second stage, data
of the partial similarity measure "S<£, m>" and the partial determination result "n<£,
m>" must be stored. The storage capacity M
1 is obtained by the following approximation:
If the following conditions are given:
then
In order to manufacture a real time speech recognition device which provides a recognition
result within 0.5 seconds after an utterance is completed, a total of "5,250,000"
computations must be completed within 2.3 seconds (= 0.5 + 120 x 0.015), provided
that the durations I and J are respectively 15 msec and the full duration from the
utterance to the response is used for computation. Thus, high speed computation of
about 0.4 µsec for each computation is required. Even if parallel processing is performed,
a large scale device is needed, resulting in high cost.
Summary of the Invention
[0011] It is, therefore, an object of the present invention to provide a continuous speech
recognition device which performs real time speech recognition and pattern matching
even in a low-speed processor and which is small in size and low in cost.
[0012] It is another object of the present invention to provide a continuous speech recognition
device which requires about half of the conventional storage capacity when total number
of computations corresponds to "J * N * I".
[0013] Assume that the reference pattern string B has concatenated Y reference patterns
B
nl, B
n2,..., B
nx-1,..., B
ny. The reference pattern string B is given as follows:
Symbol ⊕ denotes that feature vectors of each reference pattern are ordered in accordance
with the time sequence. Therefore,
According to the principle of the present invention, the reference pattern string
B given by relation (18) is optimally matched with the input pattern A given by relation
(2), in the same manner as in the conventional two-level DP algorithm, to determine
the words n "n
1, n
2, n
X-1, n
X,..., n
Y for optimal matching. Therefore, the input pattern A is determined to comprise words
"n
1, n
2,..., n
X-1, n
X,..., ny". In this case, the number Y of words is also optimally determined.
[0014] The concatenated words of the input pattern A are recognized by determining the number
of reference patterns with maximum similarities and the types of words.
Brief Description of the Drawings
[0015]
Fig. 1 graphically illustrates a computation range of a two-level DP algorithm as
a conventional continuous speech recognition means;
Fig. 2 is a graph for explaining a first step of continuous speech recognition according
to the present invention;
Fig. 3A graphically illustrates a computation range of a recursive relation with a
slope constraint when a starting point and an endpoint are fixed;
Fig. 3B is a view of an example of the computation range of the recursive relation
with a slope constraint;
Fig. 4 is a graph for explaining a decrease in total computation of continuous speech
recognition according to the present invention;
Fig. 5A is a graph for explaining the details of the first step;
Fig. 5B is an enlarged view of the cross-hatched region in Fig. 5A;
Fig. 6 is a.graph for explaining the detailed computation of the recursive relation;
Fig. 7 is a graph for explaining a second step of continuous speech recognition according
to the present invention;
Fig. 8 is a block diagram of a continuous speech recognition device according to a
first embodiment of the present invention;
Fig. 9A is a block diagram of a first recursive computation section (DPM1) of the device shown in Fig. 8;
Fig. 9B is a timing chart of control signals of the DPM- of the device shown in Fig. 8;
Fig. 10 is a block diagram of another example of a DPM1;
Fig. 11 is a block diagram of a continuous speech recognition device according to
a second embodiment of the present invention;
Figs. 12A and 12B are flowcharts for explaining the mode of operation of the continuous
speech recognition device of the second embodiment;
Fig. 13 is a flowchart for explaining Process 1 in Fig. 12A;
Fig. 14 is a flowchart for explaining Process 2 in Fig. 12A; and
Fig. 15 is a flowchart for explaining Process 3 in Fig. 12A.
Detailed Description of the Preferred Embodiments
I. General Description
[0016] In order to fully understand the present invention, the speech recognition algorithm
of the present invention is compared with a two-level DP algorithm.
[0017] The two-level DP algorithm comprises first-stage matching in which the partial similarity
measure is computed by all possible combinations of the starting points and the endpoints
to determine the partial . determination results, and second-stage matching in which
a boundary is determined for providing the maximum overall similarity measure by utilizing
dynamic programming of all possible combinations of the partial determination results.
According to the present invention, however, the overall similarity measure is not
maximized at the second stage. The maximum overall similarity measure is obtained
at the first-stage matching in the following manner. Assume that the time point "i
= p" (1 < p ≤ I) of the input pattern A is defined as the boundary between two words,
and that a maximum similarity measure "D
p = S<l, p>" is obtained from an optimum combination of the reference words and a partial
pattern "A(l, p)" of the input pattern A. The maximum similarity measure D
q between a partial pattern "A(l, q) of the input pattern A whose endpoint is a time
point "i = q" (1 < p < q ≦ I) and the optimum combination of the reference words is
given as follows:
In this case, n is stored as W , where
"S(A
(p+1, q), B
n)" indicates the similarity measure between a partial pattern "A
(p+1, q)" with the endpoint q and the reference pattern B
n of the word n. This computation is the same as that of the partial similarity measure
of the two-level DP algorithm. According to the present invention, however, the partial
similarity pressure is not independently obtained, but is obtained in the braces of
the right-hand side of relation (20). In relations (20) and (21), when the condition
"D
0 = 0" is given, D
q and W are obtained for 1 < q < I since the maximum similarity measure "D
p (p < q) is given. The maximum overall similarity measure "S<l, I> is obtained as
D
I. Thus, the first step of continuous speech recognition is completed. Thereafter,
the second step is performed in which the number Y of words which constitute a permutation/combination
"B =
Bnl ⊕
Bn2 ⊕...⊕B
nY" and their words "n
1, n
2,..., n
Y" are determined. In this procedure, a final word n
Y is obtained as W
I. However, only D. and W. are stored in order to greatly decrease the required memory
capacity and total computation of the first step. For this reason, the boundaries
of the words "n
Y, n
Y-1,...,
n2, n
1" are backtracked beginning from the time point "i = I". A recognized word W
u is output with the time point "i = I" defined as a starting point u. D
p matching is performed in the reverse direction only for the recognized word W
u. An endpoint v is obtained which maximizes the sum of D
v-1 and a similarity measure "S(A
(u,
v), B
Wu)" obtained by a backtracking partial pattern "A
(u, v)" from the starting point u to the endpoint v and the reversely concatenated reference
pattern string B as follows:
where ARGMAX is the endpoint v which gives a maximum v value in the braces of relation
(22).
[0018] The v
max is the starting point (endpoint in the reverse D matching) of the word W
u. Let the starting point u now be defined as an endpoint (starting point in reverse
D
p matching) of an immediately preceding word:
When tracking is repeated from "u
= v
max-1" to "u
= 0", all the recognized words are obtained in the reverse order. The obtained reverse
ordered word string is reversed again, so that the input word string is regarded as
recognized.
[0019] The fundamental principle of the algorithm according to the present invention is
described above. However, relation (20) cannot be computed for all p, q and n in the
first step due to the amount of total computation. If maximization at the boundary
p is first performed, relation (20) can be rewritten as follows:
Terms in the brackets of relation (24) can be substituted by a conventional dynamic
programing algorithm in which the starting point is free as "(p + 1)" whose initial
value is D
p, and the endpoint q is fixed.
[0020] The above case is described with reference to Fig. 2. If the maximum similarity measure
D
p in a duration having the time point "i = p" as the endpoint is given as the initial
value, the sum of products for maximizing the sum of the similarity measure s (i,
j) of the feature vectors a and β
j of a grid point (i, j) of a path 26 leading from a starting point 28 as (p + 1, 1)
to an endpoint 29 as (q, J
n) is obtained by the DP algorithm as "
Dp
+ S(A
(P+1, q), B
n)".
[0021] In the two-level DP algorithm, the adjustment window bounded by the lines 11 and
12 given by relation (12) and shown in Fig. 1 is arranged as a range of the (i, j)
plane for computing the similarity measure s
n(i, j) in order to eliminate wasteful computation and abrupt adjustment of the time
bases. However, according to the present invention, the adjustment window is not arranged,
but two slope constraints are included as the recursive relations in the DP algorithm.
There are various examples of slope constraints; the following is a typical example:
Initial values:
where denotes a value computable by a given processor which has the negative sign
and which has the maximum absolute value. This value is always smaller than other
values to be compared with. The following recursive relation is solved for "i = 1
to I" and "j = 1 to J":
As shown in Fig. 3B, there are three paths from different starting points to a point
31 (i, j): a path 37 from a point 32 (i - 2, j - 1) to the point (i, j) via a point
33(i - 1, j); a path 38 from a point 34 (i - 1, j - 1) to the point 31 (i, j); and
a path from a point 35 (i - 1, j - 2) to the point 31 (i, j) via a point 36 (i, j
- 1). Among these three paths, the longest path is selected. In the path 37, an increment
of 2 along the time base i of the input pattern corresponds to an increment of 1 along
the time base j of the reference pattern, so that a slope of a segment connecting
the point 34 and the point 31 is 1, while a slope of a segment connecting the point
35 and the point 31 is 2.
[0022] As shown in Fig. 3A, by utilizing recursive relation (31) for obtaining an optimal
path 40 from a starting point (1, 1) to an endpoint 46 (I, J), the search range of
the (i, j) plane is a triangular region which is bounded by a line 42 with a slope
of 1/2 at minimum and a line 41 with a slope of 2 at maximum and which is defined
by three points 45, 47 and 48. Since the endpoint 46 (1, J) is known, the search region
is restricted by a line 43 with a slope of 1/2 and a line 44 with a slope of 2. As
a result, the search region corresponds to the hatched region of a parallelogram bounded
by the lines 41, 42, 43 and 44. Thus, the recursive relation itself has a slope constraint,
so that abrupt adjustment of time bases can be prevented without arranging the adjustment
window.
[0023] Dynamic programming will be described in which an endpoint is fixed, while its starting
point is free. As shown in Fig. 2, since the endpoint 29 is fixed as (q, J
n), the similarity measures s
n(i, j) is computed in a hatched region surrounded by a line 24 with a slope of 1/2
and a line 25 with a slope of 2 by using recursive relation (31) to find an optimal
path to reach the endpoint 29.
[0024] All the points 28 (p + 1, 1) from a point 21 (q - 2·J
n, 1) to a point 22 (q - J
n/2, 1) are candidates for the starting point. If the following condition is given:
recursive relation (31) is solved from "i = k", so that
Thus, recursive relation (31) is solved for j = 1 to J
n by increasing i to q in unitary increments. The final result g (g, J
n) of the recursive relation indicates:
[0025] When the continuous speech recognition processing is performed by relation (20) with
relations (30) to (34), and when relation (31) is computed within a range of hatched
region in Fig. 2 every time the endpoint q is obtained, the overall similarity measure
between vectors and the total computation C
2 of the recursive region are given as follows:
where J is the mean value of J
n. The result is substantially the same as relation (14) but is larger than that.
[0026] If a point 50 (q, J
n) in Fig. 4 is defined as the endpoint, total computation for the region bounded by
lines 52 and 53 corresponds to 3/4·J
n2. Even if a point 51 (q + 1, J
n) which is next to the point 50 is defined as the endpoint, the total computation
for the region bounded by lines 54 and 55 corresponds to 3/4·J
n2. However, as may be apparent from the graph, the hatched region surrounded by the
lines 54 and 53 can provide the same vector similarity measure regardless of the endpoints
50 and 51. Computation of "(3/4·Jn2) * 2" substantially corresponds to (3/4·J
n2 + J
n/2). This result can be applied to every endpoint, so that the overlapped portion
can be computed once. Therefore, the total computation C
3 is given as follows:
The above relation indicates the inside of a parallelogram having as vertexes points
60 (1, 1), 61 (J
n/2, J
n), 62 (I, J
n). and 63 (I - J
n/2, 1). In this case, candidates for the starting point are all the points p' (1 <
p' < I - J
n/2) from the point 60 (1, 1) to the point 63 (I - J
n/2, 1), while candidates for the endpoint are all the points q' (Jn/2 < q' ≦ I) from
the point 61 (J
n/2, J
n) to the point 62 (I, J
n) in this dynamic programming.
[0027] Each combination of starting point and endpoint is not independently computed. All
the combinations are simultaneously computed, thus greatly decreasing the total computation.
Further, even if the parallel computation is performed, abrupt alignment of time bases
may not occur since the recursive relation includes a two-side slope constraint.
[0028] According to the dynamic programming as described above, recorsive relation (31)
is computed for the time base j (1 to J
n) of each reference pattern for all words n. The computation is continued for the
time base i of the input pattern.
[0029] Referring to Fig. 5A, assume that the computation of recursive relation (31) is completed
when "i = p". In other words, the computation of the recursive relation for the inside
region of the parallelogram having as vertexes points 60 (1, 1), 61 (J
n/2, J
n), 68 (p, J
n), and 65 (p, 1) is completed. Further, intermediate results of the recursive relation,
such as
gn(i, j) and s
n(i, j) are stored. Furthermore, the following relation:
is calculated for "i = 1 to p". For computing the relation for "i = p + 1", let the
initial value for every word be defined as D
p according to relation (33), so that:
Therefore, the following recursive relation is computed for "j = 1 to Jn" :
Therefore, g
n(p + 1, J ) is obtained for each word. n n Now, let the maximum value for n be D
p+1.
The above processing is shown in Fig. 5B. Fig. 5B is an enlarged view of the cross-hatched
region in Fig. 5A. The intermediate results D
i are included in the upper and lower sides, so that two strings are aligned. However,
these are illustrated for descriptive convenience; these are in practice the same.
[0030] Take up the word n for an example. Since the computation is completed when "i = p"
along the time axis, "g
n (i, j)" and "s
n(i, j)" at each point between D
p and (i, j) in Fig. 5B are computed for "1 < i < p" and "1
< j
< J
n". = n
[0031] According to relation (38),
[0032] Therefore, if "j = 1" is given, the recursive relation at the point (p + 1, 1) is
given as follows:
[0033] Similarly, if "j = 2" is given, the recursive relation at the point (p + 1, 2) is
given as follows:
[0034] Furthermore, if "j = J
n" is given, the recursive relation at the point (p + 1, J
n) is given as follows:
[0035] The above computation is performed for all the reference patterns, which number N.
Among the obtained results "g
1(p+1, J
1), g
2(p+1, J
2),..., g
n (p+1, J
n),..., g
N(p+1, J
N)", the maximum result D
p+1 is defined as follows:
[0036] In the above description, for each i, the total number of computations is "J * N",
so that the total computation C
3 substantially numbers "I * J * N". Further, since a memory area M
3 must store all "g
n(i, j)" and "s
n(i, j)", it is defined as follows:
[0037] The memory area M
3 is very large as indicated by relation (47). However, only "g(i - 1), j)", "g(i -
2, j)", "s(i - 1, j)", "s(i, j)" and "g(i, j)" are required to compute the ith-step
recursive relation, so that a memory area M'
3 is given in practice as follows:
[0038] Relation (48) can be rewritten for further convenience. Let "h(i, j)" be defined
as follows:
[0039] Thus, recursive relation (31) can be rewritten as follows:
or
[0040] Referring to Fig. 6, "h(i, j)" at the point (i, j) is defined by relation (49) and
is a sum of "g(i-1, j-1)" and 2 x "s(i, j)" as indicated by an arrow 85.
[0041] The first element of the maximum value of relation (31) is "{g(i-1, j-2) + 2·s(i,
j-1)}" as indicated by an arrow 86. However, according to the definition given by
relation (49), the first element becomes "h(i, j-1)". The third element of the maximum
value of relation (31) is "{g(i-2, j-1) + 2·s(i-1, j)}, as indicated by an arrow 81,
but becomes "h(i-1, j). Thus, relation (51) selects the maximum value among "{h(i,
j-1) + s(i, j)} indicated by an arrow 84, "h(i, j)" indicated by an arrow 83, and
"{h(i-1, j) + s(i, j)}" indicated by arrow 82.
[0042] According to relations (49) and (51), the used memory.areas are of three types: "h(i-1,
j)", "g(i, j)" and "h(i, j)" for j = 1 to J
n. If temporary memory n registers TEMPI, TEMP2 and TEMP3 are used, relations (49)
and (51) can be interpreted as follows:
(a) Using "TEMP1 = g(0)", "TEMP2 = h(0)" and .the initial value, repeat steps (b)
to (f) for "j = 1 to Jn".
(b) TEMP3 = h(j)
(c) h(j) = TEMPI + 2 * s(i, j)
(d) TEMPI = g(i)
(f) TEMP2 = h(j)
where h(j) is a substitute for h(i-1, j) and h(i, j), and g(j) is the same as g(i,
j).
[0043] As is apparent from the above description, used memory areas M"
3 are of two types: h(j) and g(j). Therefore,
The memory area M"
3 of the algorithm according to the present invention is smaller than the memory area
(relation (15)) of the two-level DP algorithm.
[0044] The detailed procedures of the first step or step 1 until D
i and W
i tables are prepared are as follows:
(Step 1-1)
[0045] Clear the D
i table for "i = 1 to I" by "- ∞" so that "D
0 = 0" is obtained. The complete working area of each word is defined as "- ∞".
[0046] Let
gn(j) and h
n(j) be "- ∞", for n = 1 to N, j = 1 to J
n and i = 1.
(Step 1-2)
[0047] Let the word n be 1.
(Step 1-3)
[0048] Let TEMPI be D
i-1 (= g
n(0)) and TEMP2 be
- ∞(= h
n(0)).
(Step 1-4)
[0049] Repeat step 1-5 for j = 1 to J
N.
(Step 1-5)
[0050] Let TEMP3 be h
n(j). Perform the following operations:
hn(j) = TEMPI + 2 * sn(i, j)
TEMPI = gn(j)
and
TEMP2 = hn(j) .
(Step 1-6)
[0051] If g
n(J
N) is smaller than D
i, go to step 1-7. If not, let D
i be g
n(J
N) and W
i be n.
(Step 1-7)
[0052] Let n be n + 1. If n is equal to or smaller than N, go to step (1-3).
(Step 1-8)
[0053] Let i be i + 1. If i is equal to or smaller than I, go to step (1-2).
[0054] Thus, all the intermediate results D
i and W
i are obtained by steps 1-1 to.1-8. Note that TEMPI, TEMP2 and TEMP3 are temporary
memory registers and that s
n(i, j) is the similarity measure between the input vector a. and the reference vector
of the nth word. It should also be noted that g
n(j) and h
n(j) are memory sections for storing intermediate results of the recursive relation
for each word having a length (J
N).
[0055] If the similarity computation of step 1-5 and the recursive computation are performed
without being limited by lines 68 and 69 shown in Fig. 5A, the total number of computations
is given by the following relation:
Therefore, a memory area M
4 required for the total computation is given as follows:
If the result obtained by relation (16) is substituted into relation (54),
The total computation of the algorithm according to the present invention is 1/25
that of the two-level DP algorithm, and the memory area is about 1/2 thereof.
[0056] The second step of the algorithm according to the present invention will now be described.
The intermediate results D
i and W
i are obtained in a range of 1 < i < I in the first step. The overall maximum similarity
measure D
I is given by the reference pattern string B
n whose permutation/combination B is given as follows:
The last word n
Y of relation (60) is W
I. If the boundary between the word n
Y and a word n
Y-1 immediately before the word n
Y is determined, the word immediately preceding the word n
Y-1 is readily determined from W
i. If this process is repeated up to the starting point "i = 1" of the input pattern,
the concatenated input word pattern "n
1, n
2,...., n
Y-1, n
Y" is obtained in the reverse order.
[0057] The above operation is described with reference to Fig. 7. Backtracking is performed
from the endpoint as indicated by "i = I" of the input pattern to the word W
I indicated by reference numeral 96. Assume that the boundary between the xth word
of the word string and the (x - l)th word thereof is known, and that the endpoint
(starting point in the reversed dynamic programing matching) of the (x - l)th word
is defined as "i = u". The (x - l)th word becomes W indicated by reference numeral
95. The partial similarity measure S(A
(u, v), B
Wu) is computed between the reference pattern B
Wu which is obtained by reverse ordering the reference pattern B
Wu of the word W
u in the opposite direction of "j = J
Wu to 1" and the reversed partial input pattern A(u, v) which is backtracked from the
starting point u to the endpoint v. The partial similarity is calculated by the dynamic
programming as with relation (31) above. In practice, backtracking is performed starting
from a point 91 (u, J
Wu) to a point 92 (v, 1) in a region 99 to search for a path with a maximum value. The
sum of the similarity measure S(A,
v), B
Wu) obtained upon search of the above-mentioned path and a D
v-1 indicated by reference numeral 93 is maximized. That is:
The endpoint v which maximizes the value computed in expression (61) is obtained within
the region 99 for all possible endpoints (v, 1). The obtained v is defined as v
max. The v
max is regarded as the boundary of the (x - l)th word and the (x - 2)th word. Let u be
Relations (61) and (62) are repeatedly computed until "u = 0". The recognized words
W are sequentially obtained in the reverse order. The "p", "q" and "n" in the brackets
of relation (24) are substituted by "v - 1", "u" and "W
u" in relation (61). That is:
Further, since the word n for maximizing relation (24) is determined as W
u, the result obtained by relation (61) is the same as D indicated by reference numeral
90.
[0058] In practice, when the type (e.g., symmetry) of recursive relation and computation
errors of the speech recognition device are considered, the value obtained by relation
(61) may not be the same as D
u. Therefore, the maximum value is first computed, and then v for giving the maximum
value is determined to be the boundary.
[0059] The detailed procedures of the second step or step 2 will now be described below:
(Step 2-1)
[0060] Let u and x be I and 1, respectively.
(Step 2-2)
[0061] Produce W
u as the recognized word n
x.
(Step 2-3)
[0062] Initialize the working area of dynamic programming as follows:
Furthermore, let TEMPI
= g(J
Wu + 1), and D
MAX be 0, and TEMP2 = h(J
wu + 1) be - ∞, respectively.
(Step 2-4)
(Step 2-5)
[0064] Repeat step 2-6 for "j = J
Wu to 1".
(Step 2-6)
[0065] TEMP3 = h(j)
TEMPI = g(j)
TEMP2 = h(j)
(Step 2-7)
[0066] If g (1)
+ D
i-1 is smaller than D
MAX, go to step 2-8. If D
MAX is equal to g(1) + D
i-1, V
max = 1 is obtfined. (Step 2-8)
[0067] Let i be i + 1. If i is equal to or greater than u - 2 - J
Wu, go to step 2-5.
(Step 2-9)
[0068] Since the boundary v
max of the words is obtained, let x and u be x + 1 and v
max - 1, respectively. If u is greater than zero, go to step 2-2.
(Step 2-10)
[0069] If "Y = x - 1" is given, n
x (x = 1 to Y) as the recognized word string comprising Y words of the input pattern
is concatenated in a reverse manner: "n
Y, n
Y-1,..., n
2, n
1".
[0070] Note that TEMP1, TEMP2 and TEMP3 are the same as those used in step 1, that g(j),
h(j) are the same as those used in step 1, and that D
MAX is the memory section for the maximum value of relation (61).
[0071] The total computation C
5 in step 2 is given as follows, since the word in boundary search is known:
where Y is the number of words included in the input pattern A. "Y = 4" as the mean
value of the number Y of the words and "J = 35" are substituted into relation (64)
:
The total computation C
5 in step 2 is smaller than 2% the total computation C
4 in step 1 which is given by relation (55). The total computation of the algorithm
according to the present invention is approximately given by C
4.
[0072] As described above, when feature vectors a and are similar to each other, a maximum
value is obtained as the similarity measure. However, a distance |α - β| is decreased
if they are similar to each other. Therefore, if the distance is used, the maximum
value is regarded as the minimum value, so that "- ∞" is replaced with "+ ∞".
[0073] In the first step, the maximum similarity measure is obtained by the input pattern
A and an optimal combination of the reference patterns. In the second step, utilizing
the intermediate results D
i and W
i obtained in the first step, backtracking from the endpoint of the input pattern A
is performed through a matching path obtained by the maximum similarity measure. Therefore,
the continuous speech recognition device according to the present invention can determine
the boundaries, order, and number of words in a concatenated word string.
II. Preferred Embodiments
[0074] Fig. 8 shows the overall arrangement of the continuous speech recognition device
according to a first embodiment of the present invention. An utterance is entered
through a microphone 101. An input speech signal is supplied to a feature extracting
unit 102. The frequency of the speech signal is analyzed by a Q-channel analyzing
filter so as to time sample the output level of each channel. Thus, a feature vector
α (= a
li, a
2i'..., a
Qi) is produced. The feature vector is supplied to an input pattern buffer 103 so as
to store the input pattern A for "i = 1 to I". The number I of vectors included in
the input pattern A is determined in the feature extracting section 102. Reference
numeral 104 denotes a reference pattern buffer for storing N reference patterns "
Bn (
n =
1, 2,...,
N). The reference pattern B
n(
,
,...,
) includes the Q-degree vector
= (
,
,...,
which comprises J
N reference pattern lengths. The feature vector α produced from the input pattern buffer
103 in response to a signal i and the feature n vector
produced from the reference pattern buffer 104 in response to signals j and n are
supplied to a recursive computation section 105 (to be referred to as a DPM
1105 hereinafter) in which a similarity measure s
n(i, j) between vectors is computed. With an initial value signal D
i-1, relation (51) is computed for "j = 1 to J
n". A similarity measure of each word for the input pattern A(l, i) is obtained as
g
n(i, J
n). The similarity measure g (i, J
n) from the DPM
1 105 is n n supplied to a first decision section 106 (to be referred to as a DCS
1 106 hereinafter) which executes step 1-6. The similarity measure g (i, J
n) is compared with a maximum similarity measure D
i at a time point i. If the similarity measure g
n(i, J
n) is greater than the maximum similarity measure D
i, the maximum similarity measure D
i is modified to be the similarity measure g (i, J
n). In this case, n is stored as W
i, Reference numeral 107 denotes a maximum similarity storage section (to be referred
to as a D 107 hereinafter) which stores the maximum similarity measure D
i with an endpoint of the time point i defined by relation (24). A maximum value produced
by the DCS
1 106 is stored in the D 107. A word number n which gives the maximum similarity measure
D
i produced from the DCS
1 106 is written and stored in a terminal word memory 108 (to be referred to as a W
108 hereinafter).
[0075] Reference numeral 109 denotes a DPM
2 for computing a backtracked similarity measure S(A
(u, v), B
Wu) = g(v, 1). An output g(v, 1) from the DPM
2 109 and the D
v-1 from the D 107 are supplied to a DCS
2 110 wherein a boundary point v
max for maximizing the result of relation (61) is determined and produced thereby. The
word number W
u based on data u = v
nax - 1 obtained by the boundary point v
max is stored as n
X (x = 1, 2,..., X) in an order reversing section 111 (to be referred to as an REV
111 hereinafter). The REV 111 produces the words ny (y = 1, 2,..., Y) by reversing
the time sequence. Reference numeral 112 denotes a control unit for controlling the
overall operation. The control unit 112 produces various signals which control the
feature extracting unit 102 to the REV 111 described above.
[0076] In the continuous speech recognition device with the above arrangement, in the speech
signal entered through the microphone 101, an output from the Q-channel analyzing
filter is sampled by a sampling signal t from the control unit 112 and is produced
as the Q-degree vector a = (a
1, a
2,..., a
Q). The feature extracting unit 102 supplies, to the control unit 112, a detection
signal which represents the starting point and the endpoint of the utterance and the
number I of vectors a from the beginning to the end. The input pattern buffer 103
stores the feature vector α
i from the feature extracting unit 102 in accordance with signals i (= 1, 2,..., I)
from the control unit 112. For descriptive convenience, assume that all input patterns
are stored in the input pattern buffer 103. In the control unit 112, in accordance
with step 1-1, intermediate result memory registers g
n(j) and h (j) in the DPM
1 105 and the D 107 are initialized. The control unit 112 sequentially generates signals
i (i = 1 to I). In response to each signal i, signals n are produced from 1 to N.
In response to each signal n, signals j (j = 1 to J
n) are produced wherein J
n is the pattern length of each word n.
[0077] The input pattern buffer 103 produces the feature vector α
i specified by the signal i from the control unit 112. The reference pattern buffer
104 produces the feature vector
specified by the word selection sigr,als n and j from the control unit 112. The DPM
1. 105 which receives output signals from these pattern buffers 103 and 104 updates
g
n(j) and h
n(j) by the recursive computation of relation (51). The previous vaalue of each word
which is produced by the intermediate result memory register g
n(j) or hn (j), and the similarity measure s (i, j) between the feature vectors α.
and
, are used for updating, provided that the initial value of the recursive relation
is defined as the maximum value D
i-1 at the previous unit time, that is, at (i - 1) . produced by the D 107. When "j =
J
N", the DCS
1 106 compares the similarity measure g (i, J ) with the n n maximum similarity measure
D
i having that endpoint which corresponds to the time point i until the word (n - 1).
If the similarity measure g(i, J ) is greater n than the maximum similarity measure
D
i, the similarity measure g(i, J ) is regarded as a maximum similarity n measure. At
this time, the word number n is stored as W
i in the W 108. When the above processing is completed for n = 1, 2,... N, the number
of the signals i is incremented by one. The incrementation is repeated for the number
of times corresponding to the number I of the input patterns, that is, the incrementation
is repeated for i = 1, 2,..., I, thereby obtaining all the intermediate results D
i and W. for i = 1, 2,..., I.
[0078] When the above operation is completed, the control unit 112 produces u = 1 as the
initial value. Thus, the recognized word signal W
u is read out from the W 108. The control unit 112 initializes the intermediate result
memories g(i) and h(j) in the DPM
2 109 and a maximum value detecting register D
MAX arranged in the DPM
2 to detect the maximum value of relation (61) in accordance with step 2-3. The number
of signals v is decreased one by one from u to (u - 2J
Wu) by the control unit 112. Further, in response to each signal v, the number of signals
j is decreased one by one from
JWu to 1.
[0079] A vector a is read out from the input pattern buffer 103 in response to the signal
v. A vector
is read out from the reference pattern buffer 104 in accordance with the signal j
and the word signal Wu.
[0080] The DPM
2 109 performs step 2-6 for "j = 1" using the intermediate result memory registers
g(j) and h(j) and a similarity measure s(v, j) between vectors. When "j = 1", the
DCS
2 110 compares the previous maximum value D
MAX of "v = u to (v + 1)" with the sum of the output g(v, 1) from the DPM
2 109 and the maximum similarity measure D
v-1 with the endpoint (v - 1). If the sum is greater than the maximum value D
MAX, the sum {D
v-1 + g(v, 1)} is stored as the maximum value, and the signal v is stored as v
max. The above processing until v = u -2·J
Wu. With the v
max thus obtained, an output u = v
max - 1 is supplied to the control unit 112. The control unit 112 repeats the above operation
until "u = 0". The word signal W
u sequentially obtained is stored in the REV 111 as n
x (x = 1, 2,..., Y). When "u = 0" is performed, the REV 111 produces the inverted output
n
y (y = 1, 2,..., Y) as "n
1 = n
Y, n
2 = n
Y-1,..., n
Y = n
1".
[0081] In the above embodiment, speech recognition operation is started after the input
pattern A is completely stored in the input pattern buffer 103. However, as shown
in steps 1-1 to 1-8, when one input vector a is entered, steps 1-2 to 1-7 are simultaneously
performed. The entire duration from the utterance input to the recognition result
response can be used for the speech recognition processing, thus shortening the response
time. Further, the DPM
1 105 parallel processes data of words n at high speed. The DPM
1 105 and the DPM
2 109 perform the identical operation as shown in steps 1-5 and 2-6. Since the second
step cannot be executed until the first step is completely finished, the second step
may be performed in the DPM
1. Thus, the DPM
2 may be omitted.
[0082] The.microphone 101 may be arbitrarily selected to be a telephone receiver or the
like. Furthermore, in the above embodiment, all reference numerals 101 to l12 denote
hardware. However, part or all of the processing performed by units 101 to l12 may
be performed under program control. Further, the feature extraction unit 102 comprises
a frequency analyzing filter. However, any unit can be used for extracting a parameter
which represents the speech features such as a linear prediction coefficient and a
partial correlation coefficient. The similarity measure between the vectors may be
represented by correlation, a distance, or the like.
[0083] The configuration of the DPM
1 105 which is the principle part of the continuous speech recognition device of the
first embodiment is shown in Fig. 9A. The configuration shown in Fig. 9A aims at calculating
relation (51). Reference numeral 120 denotes a similarity measure operation unit for
computing the similarity measure s (i, j) between the vectors α. and
. Reference numeral 121 denotes a temporary memory register (to be referred to as
a TEMPI 121 hereinafter) which receives g
n(i-1, j) and produces g
h(i-1, j-1). If computation is started with "j = 1" in the TEMPI 121, D
i is preset as the initial value. Reference numeral 122 denotes a temporary memory
register (to be referred to as a TEMP2 122 hereinafter) which receives h
n(i, j) and produces h
n(i, j-1). If computation is started with "j = 1", "- ∞" is preset as the initial value.
Reference numeral 123 denotes a temporary memory register (to be referred to as a
TEMP3 123 hereinafter) which temporarily stores h
n(i-1, j).
[0084] The output s
n(i, j) from the similarity measure operation unit 120 is supplied to a double multiplier
circuit 124 which produces 2·s
n(i, j). The output g
n(i-1, j-1) from the TEMPI 121 is added to the output 2·s
n(i, j) from the double multiplier 124 by an adder 125. The adder 125 produces an output
{g
n(i-1, j-1) + 2·s
n(i, j)), that is, h
n(i, j). The output h
n(i, j-1) from the TEMP2 122 and the output s
n(i, j) from the similarity measure operation unit 120 are added by an adder 126. The
adder 126 produces an output {h
n(i, j-1) + s
n(i, j)}. Furthermore, the output h (i-1, j) from the TEMP3 123 is added to the output
s
n(i, j) from the similarity measure operation unit 120 by an adder 127. The adder 127
produces an output {h
n(i-1, j) + sn (i, j)}. Reference numeral 128 denotes a maximum value detector (to
be referred to as a MAX 128 hereinafter) which selects a maximum value among the output
h
n(i, j) from a memory 130 to be described later and outputs from the adders 126 and
127. An output from the MAX 128 is supplied to a memory 129. The memory 129 stores
data as follows:
and
[0085] Data read out from the memory 129 is supplied to the TEMPI 121. The output h
n(i, j) from the adder 125 is stored in the memory 130. The storage contents are as
follows:
and
[0086] Data read out from the memory 130 is supplied to the TEMP2 122 and to the MAX 128.
Reference numeral 131 denotes a recursive control unit (to be referred to as a DPM
131 hereinafter) which controls the DPM
1 105 and the DPM
2 109. The DPM 131 supplies timing signals Tl, T2, T3, T4 and T5 to the TEMPI 121,
the TEMP2 122 and the TEMP3 123. Further, the timing signals Tl, T2,
T3, T4 and T5 are supplied to the memories 129 and 130 as write signals. The timing
signals Tl to T5 are produced respectively for each signal j in the order shown in
Fig. 9B. A timing signal TO is a preset signal and is used to preset D
i instead of g
n(i-1, 0) in the TEMPI 121 and - ∞ instead of h
n(i, 0) immediately before computation with "i = 1" is started.
[0087] In the DPM
1 105 with the above arrangement, the operation for the hatched region in Fig. 5B will
be described. All the operations for every word number n are completed up to the time
point "i = p" of the input pattern A. Therefore, D
i is obtained for "0 < i < p". Further, the memory 129 stores data g
n(p, j) (n = 1, 2,..., N; and j = 1, 2,..., J
n), and the memory 130 stores data h (p, j) (n = 1, 2,..., N; and j = 1, 2,..., J
n) .
[0088] An index j for taking out the vector
of the word number n and the reference pattern B
n is specified by the control unit 112 shown in Fig. 8. For performing the processing
for the cross-hatched region in Fig. 5B, the TEMPI 121 and the TEMP2 122 are initially
set by D
p and - ∞, respectively at time t0 in response to the timing signal T0. In synchronism
with the timing signal T3, the output h
n(1), that is, h
n(p, 1) from the memory 130 is written as "j = 1" in the TEMP3 123 at time tl. A similarity
measure s
n(p+1, 1) between vectors α
p+1 and
which are specified by "i = p + 1" and "j = 1" is computed by the similarity measure
operation unit 120. The computed result is doubled by the double multiplier circuit
124 which then produces an output 2·s
n(p+1, 1). The output 2·s
n(p+1, 1) is added to the output g
n(p, 0) = D
p from the TEMPI 121 by the adder 125. The adder 125 produces an output h
n(p+l, 1) which is written in a register h (1) of the memory 130 in synchronism with
the timing signal T5 at time t2. At time t3, the output g
n(1), that is, g
n(p, 1) from the memory 129 is written in the TEMPI 121 in synchronism with the timing
signal Tl.
[0089] An output h
n(p, 0) = - ∞ from the TEMP2 122 is added to an output s
n(p+1, 1) from the similarity measure operation unit 120 by the adder 126. An output
h
n(p, 1) from the TEMP3 123 and the the output s
n(p+1, 1) are added in the adder 127. The MAX 128 selects the maximum value among the
outputs from the adders 126 and 127 and the memory 130:
[0090] This maximum value is written as g
n(1) = g
n(p+1, 1) in the memory 129 in synchronism with the timing signal T4 at time t4. At
time t5, the output h (1) = h
n(p+1, 1) from the memory 130 is written in synchronism with timing signal T2. Thus,
the cycle of "j = 1" is completed.
[0091] The next cycle is started by data "j = 2" from the control unit 112. At time t6,
an output h
n(2), that is, h
n(p, 2) from the memory 130 is written in the TEMP3 123. At time t7, the output g
n(p, 1) from the TEMPI 121 and the output 2·s (p+1, 2) from the douple multiplier circuit
124 is written as h
n(2) in the memory 130. At time t8, g
n(2) = g
n(p, 2) is then written in the memory 129. At time t9, the maximum value among the
output h (2) = h
n(p+1, 2) from the memory 130, the output {h
n(p+1, 1) + s
n(p+1, 2)} from the adder 126, and the output {h
n(p, 2) + s (p+1, 2)} from the adder 127 is written as g
n(2) = g
n(p+1, 2) in the memory 129. At time t10, the output h
n(2) = h
n(p+1, 2) from the memory 130 is written in the TEMP2 122. Thus, the cycle of "j =
2" is completed. When the above cycle is repeated until j = J , the output g
n(J
n) from the memory 129 becomes g
n(p+1, J
n). This output is supplied to the DSC
1 106 in Fig. 8.
[0092] The above processing is performed by relation (51) which is the modified relation
(31). In addition to relation (51), there exists, for example, the following recursive
relation with slope constraints:
[0093] The slopes of relation (70) are 2/3, 1, and 2, which indicates that the input pattern
length may be changed with +50% of the reference pattern length. The allowable pattern
length range of relation (70) is narrower than that of relation (31) which has a range
of -50% to +100% for slopes 1/2, 1 and 2. In the same manner that relation (31) is
modified to obtain relation(51), let f
n(i, j) be defined as follows:
Relation (70) can be modified as follows:
Or
Step 1 is thus modified as follows: (Step 1'-1)
[0094] Let's define:
D0 = 0
Di = - ∞ for i = 1 to I
[0095] Further, let i be 1.
(Step l'-2)
(Step l'-3)
[0097] Let's define:
TEMPI = Di-1 (= gn(0)) and
TEMP2 = - ∞ (= hn(0))
(Step 1'-4)
[0098] Repeat step 1'-5 for j = 1 to J
n, (Step 1'-5)
TEMP3 = hn(j)
TEMP1 = gn(j)
TEMP2 = hn(j)
(Step l'-6)
[0099] If g
n(J
n) is smaller than D
i, go to step 1'-7. If not, let D
i and W
i be g
n(Jn) and n, respectively. (Step 1'-7)
[0100] Let n be n +1. If n is equal to or smaller than N, go to step 1'-3.
(Step l'-8)
[0101] Let i be i + 1. If i is equal to or smaller than I, go to step l'-2.
The above program sequence is performed by the circuit shown in Fig. 10. The circuit
is operated at the same timings as in Fig. 9B. The reference numerals used in Fig.
9B denote the same parts in Fig. 10, and a detailed description thereof will be omitted.
[0102] Examples of recursive relations without slope constraints are as follows:
Or
The second example is described in Patent Disclosure DE-A 26 10 439 . Since neither
recursive relation has a slope constraint, an aligning window as bounded by lines
11 and 12 shown in Fig. 1 is required for avoiding the abrupt alignment of the time
bases.
[0103] In both above recursive relations, the abrupt alignment occurs locally. According
to experiments in speech recognition, unsatisfactory results have been reported.
[0104] Fig. 11 shows a continuous speech recognition device according to a second embodiment
of the present invention. An utterance entered as an analog speech signal through
a microphone 101 is converted to a digital signal in an A/D converter 141. Reference
numeral 142 denotes a data memory for storing data of the input pattern A, the reference
pattern B, the intermediate results g
n(j), h
n(j), g(j) and h(j), the maximum similarity measure D
i, and the terminal word W
i. Reference numeral 143 denotes a program memory. The speech signal converted to a
digital signal is coupled to a CPU 144. The program from the program memory 143 is
executed in the CPU 144.
[0105] In order to describe the above processing in detail, the speech signal entered from
the microphone 101 is converted to a digital signal by the A/D converter 141. The
digital signal is fetched in the CPU 144 with a predetermined time interval, for example,
100 p sec.and is then stored in the data memory 142. When 150 digital signals are
written in the data memory 142, the CPU 144 performs fast Fourier transform (FFT)
to obtain the power spectrum which is multiplied by 16 triangular windows. Thus, the
same result as in a 16-channel frequency analyzing bandpass filter are obtained. The
result is defined as the input vector a. One hundred and fifty pieces of data are
obtained every 15 msec. The time interval of 15 msec. is defined as one frame.
[0106] The mode of operation of the CPU 144 by the program stored in the program memory
143 will be described with reference to flowcharts in Figs. 12 to 15.
[0107] A variable i
1 used in the flowcharts is an index representing an address for storing the vector
a computed in the interrupt processing. A variable ℓ is a counter for counting low-power
frames for detecting a terminal end in an interrupt loop. A variable I indicates the
number of vectors a from the starting point to the endpoint. A variable i
2 is an index for reading out the input vector a in the speech recognition processing.
In the low-power frame during the continuous word pattern, process 2 (corresponding
to step 1) is not performed, and the flow advances. A variable i3 is an index for
reading out the input vector a for executing process 2. D
i, g (j), h
n(j)' W
i, g(j) and h(j) are data stored in the data memory 142; D
i is the maximum similarity measure with the ith frame as an endpoint; g
n(j) and h
n(j) are registers for intermediate results of the recursive relation for the word
having the word number n in process 2; W
i is the terminal word of the word string for giving the maximum similarity measure
D
i; and g(j) and h(j) are registers for storing the intermediate results of the recursive
relation in process 3 (corresponding to step 2). A variable j is an index for reading
out the vector β
j of the reference pattern. A variable n indicates a word number. A constant J indicates
a time duration (the number of frames) of the word having the word number n. A constant
N is the number of reference patterns. Variables TEMPI, TEMP2 and TEMP3 are temporary
memory registers of the DPM
1 105. A variable u is an index for providing a starting point of a partial pattern
of the reverse pattern matching in process 3. A variable v is an index for providing
an endpoint of the partial pattern of the reverse pattern matching. D
MAX is a register for storing a detected maximum value given by relation (61). v
max is a register for storing the index v which gives D
MAX. A variable x is an index for storing a recognized word number n
x. s (i, j) is the similarity measure between the vectors α
i and
. Symbol - indicates the maximum negative value obtained in the CPU 144.
[0108] The main program starts with start step 200 as shown in Fig. 12A. In step 201, a
flag is initially set to be low so as to indicate detection of the starting point
and endpoint of the utterance. Thus, the interrupt from the A/D converter 141 is enabled
every 100 p sec. In the following steps, the interrupt processing for data fetching,
computation of the feature vector a, and detection of the starting point and the endpoint
are performed in parallel with the speech recognition processing.
[0109] Interrupt processing steps 220 to 233 are first described with reference to Fig.
12B. When the interrupt occurs, interrupt process step 220 is initiated. In step 221,
digital data from the A/D converter 141 is stored in the data memory 142. It is determined
in step 222 whether or not the number of data has reached 150. If NO in step 222,
the interrupt process is terminated in return step 223. When 150 pieces of data are
written in the CPU 144, the computation of the vector a is performed in step 223.
It is then checked in step 224 if the starting point detection flag is set to logical
level "0". If YES in step 224, it is determined in step 127 whether or not 16 the
sum of powers (e.g., the sum E a of the elements of the vector a) is higher than a
threshold value. If NO in step 227, the interrupt process is interrupted in step return
step 233. However, if YES in step 227, it is determined that the starting point is
detected. In step 228, the starting point flag is set to logical level "1" and the
index i
1 is set to logical level "1"; vector α
il is defined as a
1 and is stored in the input pattern buffer A. In step 229, the counter 1 is set to
logical level "0", so that in return step 223 the interrupt process is interrupted.
[0110] Meanwhile, when the starting point detecting flag is set to logical level "1" in
step 224, the index i
1 is increased by one in step 225 and is defined as the input vector α
il and is stored in the input pattern buffer A. In decision step 226, if the power of
the input vector is higher than the threshold value, the flow advances to step 229.
Otherwise, the input vector is regarded as a low-power frame, and the counter ℓ is
increased by one.
[0111] In step 231 which represents a decision box, it is checked whether or not the count
of the counter ℓ has reached 20, that is, 20 low-power frames have been continuously
counted. If NO in step 231, the flow advances to return step 223. Otherwise, it is
determined that the input utterance is completed, so that in step 232 the number of
effective vectors a from the starting point to the endpoint is defined as I
1, and the endpoint detecting flag is set to logical level "1". An interrupt from the
A/D converter 141 is prohibited, and in step 233 the interrupt process is interrupted.
[0112] By the above interrupt process, the vectors a are fetched in the input pattern buffer
A every 15 msec.
[0113] The steps after step 202 in the main program will be described below. In process
1 (steps 240 to 245 in Fig. 13) in step 202, initialization corresponding to step
1-1 is performed.
[0114] In decision step 203, it is waited until the starting point detecting flag is set
to logical level "1". When the flag is set to logical level "1", a speech input is
regarded as started. In step 204, the indexes i
2 and i3 are initialized to be logical level "1". In decision step 205, indexes i
1 and i
2 which are used for the interrupt process are compared. If the index i
2 is equal to or smaller than i
1, the flow advances to step 206 which represents a decision box. If the power of the
vectors α
i2 is smaller than the threshold value, i
2 is regarded as a low-power frame during the speech input. In step 207, the index
i
2 is increased by 1. Subsequently, in decision step 208, the logical status of the
endpoint detecting flag is checked. If the flag is set to logical level "0", the endpoint
is regarded as not being detected. Then, the flow returns to step 205.
[0115] However, in step 206, if the power of the vectors is higher than the threshold value,
the flow advances to step 212 in which the index i
2 is increased by one. Process 2 which corresponds to steps 1-2 to 1-7 is performed
in step 213. In step 214, the index i3 is increased by one. In step 215, the indexes
i3 and i
2 are compared. If the index i3 is smaller than the index i
21 the flow returns to step 213 and process 2 is continued. However, if the index i3
is equal to or greater than the index i
2, the flow returns to step 205.
[0116] If the endpoint detecting flag is set to logical level "1" in step 208, the indexes
i
3 and i
2 are compared in step 209. If process 2 in step 213 is not completed within 15 msec.,
fetching of vectors a may be performed prior to completion thereof. Therefore, if
the endpoint is detected, an unevaluated input vector a may be present in the CPU
144. Thus, the indexes i
3 and i
2 are compared. If the index i
3 is equal to or smaller than the number I
1 of input vectors, the same operation as in steps 213 and 214 is performed in steps
210 and 211. However, in step 209, if it is determined that the index i3 is greater
than I
1, all the input vectors are regarded as evaluated, and process 3 (correspond to step
2) in step 216 is then performed to obtain the recognized word string n
x. In step 217, the order of the word string n
x is reversed to obtain the reversed word string n
y. Thus, the continuous speech recognition process is completed.
[0117] The detail of process 2 is illustrated in Fig. 14. Step 251 corresponds to step 1-2;
step 252 corresponds to step 1-3; steps 253, 256 and 257 correspond to step 1-4; steps
254 and 255 correspond to step 1-5; steps 258 and 259 correspond to step 1-6; and
step 261 corresponds to step 1-7.
[0118] Fig. 15 shows the detail of process 3. Step 271 corresponds to step 2-1; step 272
corresponds to step 2-2; steps 273 and 274 corresponds to step 2-3; and step 275 correspond
to step 2-4. Symbol i used in the second step is replaced with v in the flowchart
in Fig. 15. Steps 276, 279 and 280 correspond to step 2-5; steps 277 and 278 correspond
to the recursive computation in step 2-6; steps 281 and 282 correspond to the maximum
value detection in step 2-7; steps 283 and 284 correspond to step 2-8; step 285 corresponds
to step 2-9; and step 286 corresponds to step 2-10.
[0119] In the second embodiment, process 3 in step 216 which corresponds to step 2 is performed
such that the endpoint is detected when 20 low-power frames are continuously counted.
However, the detection may be performed simultaneously when the first low-power frame
is detected. If an effective power frame is entered before the endpoint is detected,
the results by process 3 become invalid. With the above arrangement, the result n
(or n ) is known when the endpoint is detected, thus requiring shorter response time.
[0120] According to the experimental results of the continuous speech recognition process
according to the continuous speech recognition device of the present invention, for
a total of 160 numbers (40 numbers for each of 2 to 5 digits), 96.3% of numerals were
recognized. For a total of 560 words (each digit of the number being defined as one
word), 99.2% of words were recognized. The above figures correspond to the result
in which 99.5% of 1000 separately occurring discrete words were recognized. Thus,
the continuous speech recognition means according to the present invention is very
effective.
[0121] It is understood from the above description that the following advantages are provided
according to the continuous speech recognition device of the present invention:
(1) In order to obtain the maximum similarity measure D between the partial pattern
A(l, q) (of the input pattern A) which has as its starting point "i = 1" and as its
endpoint "i = q", and a suitable combination of the reference patterns, the partial
pattern A(l, q) is divided into sub-partial patterns A(l, p) and A(p+l, q). The maximum
similarity measure Dp of A(l, p) is added to the similarity measure S(A(p+1, q), Bn) obtained by matching between A(p+l, q) and the reference pattern Bn. The maximum value of the sum with respect to p is obtained by the dynamic programing
algorithm. Further, a means for determining the maximum value Dq with respect to n only once performs the computation of the similarity measure s(αi,
) between vectors with respect to a combination of i, j and n. Therefore, the number
of total computations in the algorithm of the present invention is about 1/25 the
number in the conventional algorithm.
(2) When one input vector a (I ≤ q < I) is entered, all the words with word number
n (1 to N) and time bases j (1 to Jn) of each word can be computed to obtain all the D and Wq, so that 98% of the total computation in the first step is parallel processed upon
entry of the speech input. Therefore, the time duration from utterance to recognition
response can be effectively used.
(3) According to the table of the maximum similarity measure Di and the end word Wi (1 ≤ i < I) which are obtained in the first step, the endpoint (i = I) of the input
pattern is defined as the first boundary. For these words determined only by the boundaries,
backtracking is performed. Therefore, the boundary of the immediately previous word
can be readily obtained by the dynamic programing algorithm.
(4) In the dynamic programing algorithm in which the boundary is obtained in the second
step, since one of the two words at the boundary (endpoint of the input pattern) is
determined, the total amount of computations is very small.
(5) Since it is possible to complete computation of the first step at the same time
as the utterance is completed, and since the total number of computations in the second
step is as small as 2% of that in the first step, response can be made as a whole
at the same time as the utterance is completed.
(6) In the pattern matching device having intermediate result memory registers gn(J), hn(j) and so on and TEMP1, TEMP2 and TEMP3, the total memory area is small as indicated
by relation (52) or (54) compared with the case in which part or all of gn(i, j) and sn(i, j) are stored as indicated by relation (47) or (48). Pattern matching can be readily
performed with hardware.
(7) Since the amount of data processing of the algorithm of the present invention
is 1/25 that of the conventional algorithm, low-speed elements can be used, resulting
in low lost.
(8) If the same elements used in the conventional device are used for the device of
the present invention, the number of reference patterns can be increased by 25 times.
Therefore, the type of recognized words can be increased.
(9) Since the memory area of the device of the present invention is half of that of
the conventional device, the device as a whole is low in cost and small in size.
(10) Although parallel processing can be considered for high speed processing, the
device of the present invention can perform the total computation during the time
taken for one vector input in the conventional device, the total number of computations
being {(2 * r + 1)* N}, and the total memory area being {(2 * r + 1)2*N*2+ M1} . Further, if this device is used for parallel processing, the total number of computations
is 1/25 that of the conventional device, . and the total memory area is about 1/17
that of the conventional device.
(11) The number Q of degrees of the feature vector, and the number of vectors of the
input and reference patterns per unit time are generally increased to improve speech
recognition. If the algorithm of the present invention is applied to the conventional
device, and if the number of recognized words and the response time are the same as
in the conventional case, the product of the number of vectors per unit time and the
number Q of degrees of the vector can be increased by 25 times. Thus, higher recognition
efficiency can be expected.
(12) Since a recursive relation with a two-side slope constraint is used, no abrupt
adjustment of the time bases occurs even if the dynamic programing algorithm is simultaneously
performed for the plurality of starting points and endpoints. Thus, the total amount
of computations is greatly decreased.
(13) The maximum similarity measure D , between the partial pattern A(l, p) of the
input pattern having as endpoint p the immediately previous frame of each starting
point (p + 1) and a combination of proper reference patterns, is used as an initial
point of each of a plurality of starting points. A combination of reference patterns
which optimally approximates the partial pattern A(l, q) having a given endpoint q
is obtained by, summation of the initial value D with similarity measures S(A(p+1, q), Bn) of the partial pattern A(p+l, q) and each reference pattern Bn. Thus, continuous speech recognition can be performed with a substantially same amount
of processing as in the case of discrete word recognition.